[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: 950822: Solaris 2.4 installation (netCDF)

This archive contains answers to questions sent to Unidata support through mid-2025. Note that the archive is no longer being updated. We provide the archive for reference; many of the answers presented here remain technically correct, even if somewhat outdated. For the most up-to-date information on the use of NSF Unidata software and data services, please consult the Software Documentation first.

  • Subject: Re: 950822: Solaris 2.4 installation (netCDF)
  • Date: Tue, 22 Aug 1995 20:32:36 -0600
Joel,

> I followed your suggestions and was able to make "all", 
> "test", and "install" successfully.  In order to fix
> things I changed my PATH so that 'which cc' gave a
> response of /opt/SUNWspro/cc rather than /usr/ucb/cc.
> The former should be the correct path to access the compiler
> running on my machine (SPARCcompiler C 3.0.1), which should (?)
> be compatible with ANSI C.  It was also necessary to add
> /usr/ccs/bin to the end of the PATH so that 'ar' and 'make'
> could be found.
> 
> Two possible problems with the installation were observed:
> 
> 1.  During the configure, "checking for the type of operating
>     system" seemed to take a very long time, especially compared
>     to the other checks that were performed.  Is this normal ?

I think it's normal.  The configure script builds a little C program with a
bunch of #ifdefs to test for various operating systems, then compiles it to
find out which built-in macros were defined.  Since this is the first time a
C compiler is used by the script, it's possible it has to auto-mount the disk
on which the C compiler resides, check with the floating license manager to
make sure you aren't exceeding the number of licenses available for your C
compiler, possibly wait for a license if it's in use, ...  I imagine if you
ran the configure script again right after running it the first time, this
part would ruin quickly the second time.

> 2.  During the make of "all", a number of "warning: semantics of
>     ">" change in ANSI C; use explicit cast" messages were displayed.
>     None of these proved fatal, however, and the subsequent making of
>     "test" and "install" proceeded without any problems.  (The log
>     for the make of "all" is attached to this mail message.)
>     Is there an easy way to get things to compile without the 
>     warnings ?  Or can these warnings be safely ignored ?

We've fixed the source for these messages in the next release, but they're
harmless and can be ignored.  I think there is some sort of compiler flag
such as -Xs or -Xa you can use to avoid these warnings.

______________________________________________________________________________

Russ Rew                                           UCAR Unidata Program
address@hidden                              http://www.unidata.ucar.edu